CHANGES IN THE BUILDING STRUCTURE OF CHOROSZCZ AND NAREW DISTRICTS OVER THE YEARS 1931–1998

Paweł Próchnicki
Katedra Ochrony i Kształtowania Środowiska Politechnika Białostocka

ABSTRACT

Demographic and economic development affects the natural environment in various ways. One of the elements of that influence is undoubtedly the development of town and rural settlement. The objective of the work was to analyze the changes in spatial building structure of two rural districts: Choroszcz (directly neighbouring the city of Białystok) and Narew. Structure changes in years 1931, 1953 and 1998 were compared. In the study, GIS methodology and spatial statistics were applied. Analyses showed a strong influence of development of the suburban zone on the development of rural areas directly neighbouring the city. That influence is visible both in the spatial aspect and in the building structure.

Key words: settlement, spatial structure, diachronic, rural district, spatial statistic, Moran’s I, Ripley’s K.

INTRODUCTION

In many regions of the world, urbanization strongly affects natural environments and transforms landscape, influencing the structure, function and dynamics of ecosystems [15]. Transformations of terrain, combined with building expansion, significantly affect biodiversity, energy flow, biochemical cycles and climate both on the regional and local scale [2, 4, 16, 22]. In the context of spatial processes transforming the landscape, urbanization is classified as a process of perforation or abrasion [10]. Due to a great escalation of urbanization processes, it is important to monitor and analyze the mechanisms of those processes. That will allow for the development of new standards in spatial management and planning and for retaining ecological balance, minimizing negative impact on the environment [4, 10]. Observation of urbanization processes, as well as of changes in their spatial and time dynamics, may be carried out on the basis of analysis of aerial, archival and contemporary geographical and topographical maps and bibliography [10]. For the analysis of the data gathered, GIS methodology and spatial statistics, allowing for calculation of measurable statistical measures, are used for comparative analyses [15, 21].

In geographical studies, more and more often so-called spatial statistics is used [11, 18]. The basis of its application is consideration in calculations not only of statistical measures of qualitative and quantitative variables features, but also of their spatial location. Special software, directly using numerical maps GIS may be helpful in calculations [18, 20]. With the use of information on spatial location of phenomena and objects, a number of tests can be conducted (e.g. Ripley's K), assessing the spatial structure of point data of one category [6, 12, 19] and of many categories [1, 3, 7]. Numerous tests and analyses (e.g. Moran's I, Geary's c, LISA), diagnosing the type of spatial distribution, are based on spatial autocorrelation [13, 18]. Other correlation tests, i.e. Mantel test and Pearson's test, compare the spatial distribution of two variables and constitute the basis for analyses of spatio-tempolar variability [8, 9, 11].  

The objective of the work was to analyze changes in building structure of two rural districts, Choroszcz and Narew. Choroszcz District is directly adjacent to the city of Białystok, while Narew District is located approximately 34 km away from the city. The analysis was diachronic in character; changes in the structure in years 1931, 1953 and 1998 were compared. Evaluation of the influence of the city and its developing suburban zone may have important implications on space management and the assessment of impact on the natural environment.

MATERIALS AND METHODS

In the study, archival topographic maps in the scale 1:100 000 from the years: 1931, 1953 and 1998 were used. On the basis of those maps, numerical maps of buildings distribution of Choroszcz and Narew Districts were created, with the help of GIS software – Geomedia Professional. Bitmap pictures of topographical maps were placed in a common, homogenous coordination system. Borders of properties and individual buildings were transferred to the numerical maps; their areas were measured and saved as the objects' attributes. For statistical calculations, the following were prepared:

• choropleth maps with the field size 500x500 m, the housing area in each field was summed up;
• centroids of objects representing particular properties or compact housing; information on the area of the represented object on the map was linked to them.

In calculations of spatial statistical measures, the software: Passage 2 [20] and SAM 4 [18] was used.

Pearson's test is a very popular correlation test, used in many disciplines. It determines the level of linear correlation between variables [14]. Testing spatial data is possible thanks to the introduction of CRH procedure [5]. That procedure uses the similarity between variance and degrees of freedom correction and makes it possible to establish the significance level of the test. It can be used to study the similarity in point data distribution in two matrices; geographical coordinates are then treated as variables [8, 11]. Pearson's r coefficient may assume values between -1 and +1 [18]. That test was used to analyze the similarity of housing distribution in successive years.

Ripley's K test allows for evaluating spatial distribution of point data by comparing them to a hypothetical, ideally random sample. Data distribution is compared to Poisson distribution. The test analyzes variability of dispersion (concentration) coefficient as the function of distance. The result of the test is a graph L(t) as the function of distance. Values L(t)<0 are interpreted as regular distribution, L(t)=0 as random distribution and L(t)>0 as cluster distribution [6, 19].

Moran's I test is diagnostic for the spatial structure of constant data in the fields of a choropleth map. It is an autocorrelation test, studying spatial dependencies (correlations) of objects of the same group. Coefficient I is calculated as iteration for successive distance ranges. The test is standardized and assumes values from -1 to +1. The result of the test is a variogram presenting variability of autocorrelation as the function of distance. The course of variogram line, particularly the point of crossing with the X axis, allows for determination of spatial distribution type: random, gradient, even or cluster (regular or irregular) [13, 11, 18].

STUDY AREA

Two districts in Podlaskie Voivodship: Choroszcz and Narew (Fig. 1) were selected for the study. Choroszcz District is directly adjacent to the city of Białystok, while Narew District is located approximately 34 km (straight) away from the city. According to data for the year 2004: the area of Choroszcz District is 163.5 km² and Narew District – 334.5 km². Choroszcz District is inhabited by 12,700 residents and Narew District – by 4,300 residents. Those numbers give 77.7 and 12.7 persons/km², respectively. Both districts have an agricultural character: in Choroszcz District, 61% of the area is farmlands and 17%, woodlands; in Narew District, 25% is farmlands and 65%, woodlands.

RESULTS AND DISCUSSION

The area of housing in Choroszcz District in 1931 amounted to 1.41 km², in 1953 – 2.92 km² and in 1998 increased to 3.06 km². In the study period, an increase of over 100% was noted. That increase results from the development of the area of towns and villages developing spatially, as well as from dispersed housing in the Eastern part of the district, close to the border with the city of Białystok (Fig. 2).

The area of housing in Narew District in 1931 amounted to 1.23 km², in 1953 – 3.2 km² and in 1998 increased to 4.11 km². In the study period, an increase of over 340% was noted. In that district, the development of housing was the most intensive in the post-war period until the year 1953 and was determined predominantly by the spatial development of villages. In the next years, until 1998, it was much slower but still the cluster system was retained (Fig. 3).

Spatial distribution of buildings (based on the map of centroids of individual properties) in successive time pairs was analyzed with the Pearson's test. Pearson's correlation coefficient r in Choroszcz District for the pair 1931-1953 amounted to 0.85, and for the next time pair 1953-1998, it amounted to 0.78. High similarity to the first time pair is characteristic of development of rural areas: housing develops around constant points (villages) and is conditioned by historical assumptions; no other settlement points are created. A decrease of similarity (by approximately 0.07) in the next time pair is a result of the development of dispersed settlement at the border with the Białystok city. Around cities, suburban zones usually develop in the form of concentric stripes entering rural areas and modifying their structures and functions [10, 22].

Pearson's correlation coefficient r in Narew District for the pair 1931-1953 amounted to 0.82, and for the next time pair 1953-1998, it amounted to 0.84. Similarly to Choroszcz District, a high correlation for the first time pair is a result of towns and villages development. In the next years, the tendency of cluster housing development was retained in that district, resulting in enhancing correlation for the next time pair.

Another test conducted was Ripley's K test, performed, similarly to Pearson's test, on the map of centroids (Fig. 4). In both districts, L(t) values were higher than 0, meaning that settlement has predominantly cluster character. In both districts, concentration strength is on a similar level; L(t) reaches the maximum value of approximately 1000 (Fig. 4). In both districts, housing clusters in 1931 and 1953 were of a similar character. In Narew District, the concentration strength in close distances is much greater. A weaker character of that phenomenon in Choroszcz District in the year 1953 probably results from the increasing impact of the suburban zone. In 1998, in Narew District, the tendency to create clusters was retained. As for Choroszcz District, a clear change of trend is seen. Ripley's K test shows significant growth of clusters – L(t) at the distance of about 50 km reaches its maximum value – approximately 1050. That result of the test is a consequence of the Białystok suburban zone growing from the city side in the form of dispersed individual houses. The tendencies described are also observable at the visual assessment of source maps (Fig. 2 and 3).

The housing structure was also analyzed with the Moran's I spatial autocorrelation test (Fig. 5). Variograms were interpreted in accordance with methodology suggested by [11]. The variogram line for the years 1931 and 1953 for both districts is similar in character: a few times it crosses the zero X axis, and positive and negative tilts are quite regular and reach similar levels. It enables us to describe the spatial housing structure in the years 1931 and 1953 in both districts as clusters of irregular size and distribution. That structure type is typical for rural areas [10, 16]. We can characterize the system in Narew District in the year 1998 in a similar way. A change in the character of variogram is visible in position for the year 1998 in Choroszcz District. In Choroszcz District, the beginning of the variogram line rises (up to the I value of about 0.1) and drops gently (at the distance of approximately 12 km). Such a variogram indicates the development a gradient structure. Emerging of such a structure is a result of unilateral entering of the suburban zone. Rural areas adjacent to the city are an interesting place of investment for the development of housing and service centres. Transformation of farmland and using it for building reaches farther and farther away from cities and promotes the development of gradient structure [10]. That effect is not observable in the case of Narew District.

CONCLUSIONS

In the light of analyses of cartographic material and spatial analyses on the territory of the districts under consideration, changes in building structure are visible. The spatial development of settlement results from the impact of demographic and economic factors. Choroszcz District is under a strong pressure of urbanization processes arising out of the development of Białystok suburban zone. Until 1953, settlement in that district had been developing in a way typical for rural areas, with cluster systems. After 1953, a clear tendency is seen to change that structure into a gradient system. Intensification of housing development is the strongest near the borders with the city. In Narew District, the development of dispersed housing was not observed. On its territory, the spatial building system retains a constant cluster structure connected with the developing grid of towns and villages.

REFERENCES

1. Andersen M.; 1992. Spatial analysis of two species interactions. Oecologia; 91; 134-140.
2. Baker L.A., Hope D., Xu Y., Edmonds J. and Lauver L. 2001.; Nitrogen Balance for the Central Arizona-Phoenix (CAP) Ecosystem. Ecosystems; 4; 582-602.
3. Barot S., Gignoux J., Menaut J.C. 1999.; Demography of savanna palm tree: predictions from comprehensive spatial pattern analysis. Ecology; 80; 1987-2005.
4. Breuste J., Feldmann H. and Uhlmann O. 1998.; Urban Ecology; Berlin; Springer.
5. Clifford P., Richardson S., Hémon D. 1989.; Assessing the significance of the correlation between two spatial processes. Biometrics; 45; 123-134.
6. Dale M.R.T. 1999.; Spatial pattern analysis in plant ecology. Cambridge; Cambridge University Press.
7. Dixon P.M. 2002.; Nearest-neighbor contingency table analysis of spatial segregation for several species. Ecoscience; 9; 142-142.
8. Dutilleul P. 1993.; Modifying the t test for assessing the correlation between two spatial processes. Biometrics; 49; 305-314.
9. Dutilleul P., Stockwell J.D., Frigon D., Legendre P.. 2000.; The Mantel test versus Pearson's correlation analysis: Assessment of the differences for biological and environmental studies. Journal of Agricultural, Biological and Environmental Statistics; 5; 131-150.
10. Forman R. 1995.; Land mosaics, the ecology of landscapes and regions. Cambridge; Cambridge University Press.
11. Fortin M.-J., Dale M.R.T. 2005.; Spatial analysis: a guide for ecologists. Cambridge; Cambridge University Press.
12. Haase P. 1995.; Spatial pattern analysis in ecology based on Ripley's K-function: introduction and methods of edge correction. Journal of Vegetation Science; 6; 575-582.
13. Legendre P., Fortin M.-J. 1989.; Spatial pattern and ecological analysis. Vegetatio; 80; 107-138.
14. Łomnicki A. 2000., Wprowadzenie do statystyki dla przyrodników. [Introduction to statistic for environmentalist]. Warszawa; PWN. [in Polish]
15. Luck M., Wu J. 2002.; A gradient analysis of urban landscape pattern: a case study from the Phoenix metropolitan region, Arizona, USA. Landscape Ecology; 17; 327-339.
16. McDonnell M.J., Pickett S., Groffman P. and Bohlen P. 1997.; Ecosystem processes along an urban-to-rural gradient. Urban Ecosys.; 1; 21-36.
17. Rangel T.F., Diniz-Filho J.A.F., Bini L.M. 2010.; SAM: a comprehensive application for Spatial Analysis in Macroecology (Version 4). Ecography; 33; 46-50.
18. Ripley B.D. 1976.; The second-order analysis of stationary point processes. Journal of Applied Probability; 13; 255-266.
19. Rosenberg M.S., Anderson C.D. (in press) 2011.; PASSaGE: Pattern Analysis, Spatial Statistics and Geographic Exegesis Version 2. Methods in Ecology and Evolution 2.
20. Solon J. 2009.; Spatial context of urbanization: Landscape pattern and changes between 1950 and 1990 in the Warsaw metropolitan area, Poland. Landscape and Urban Planning; 93; 250-261.
21. Sukopp H. 1998. Urban ecology-Scientific and practical aspects. Breuste J. (ed)., Urban Ecology; Berlin; Springer; 3-16.

Accepted for print: 30.09.2011

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.